JIA-2019-11

2623 WANG Shi-chao et al. Journal of Integrative Agriculture 2019, 18(11): 2619–2627 (Fig. 4-B). Correlation analysis indicated the CPISP value was not significantly correlated with the experimental years at the 5% level of significance ( R 2 =0.02, P =0.58). The LTEs clearly revealed the CPISP trend, showing that CPISP decreased very slowly at a rate of 0.07% per year. The CPISP values of rice-rice rotation in southern China over the 30 years of the experiment are shown in Fig. 4-C. CPISP changes under the rice-rice rotation system ranged from 33.5 to 81.9%. The regression relationship between the CPISP for early rice (y) and experimental years ( x ) was y =59.1+27.8e –0.57 x ( R 2 =0.15, P <0.05). According to this equation, the CPISP decreased sharply at an average rate of 3.1% per year for the first two years and then remained steady during the following years. Differences in intercropping systems The CPISP values revealed the significant response of different cropping systems (Fig. 5). In a mono-cropping system, the CPISP of wheat-maize-soybean rotation ((65.9±16.1)%) was significantly higher ( P <0.05) than that for the continuous maize ((43.5±11.8)%) and wheat-maize systems ((44.5±19.2)%), whereas no significant difference was found between wheat-maize rotation in Northwest China and continuous maize systems in Northeast China. In double-cropping systems, the CPISP of paddy fields (early-late rice system) was significantly higher ( P <0.05) than those in drylands (wheat-maize rotation) and paddy- upland rotation (wheat-rice system) by 55.5 and 21.4%, respectively. Overall, the wheat-maize-soybean rotation in Northeast China had the highest CPISP among all of the cropping systems, followed by paddy fields in the south of China, paddy-uplands in northern China, wheat-maize in Northwest China, continuous maize in Northeast China and upland in northern China. 4. Discussion 4.1. Contribution percentage of inherent soil pro- ductivity In our study, CPISP3 was the lowest in the 1980s, the highest in the 1990s, and at an intermediate level in the 2000s (Fig. 1). This result occurred because total nutrient levels increased at the studied sites between the 1980s and 1990s (Table 2). In particular, the input of chemical fertilizer increased rapidly, while the input of organic fertilizer was maintained (Appendix B). Consequently, soil nutrient content (as reflected in the available phosphorus (AP) and available potassium (AK) values) increased (Table 2) in the 0–20 cm soil depth. In the 1990s, soil nutrient content was not the main factor affecting crop growth, and then fertility improved. This phenomenon explains the increase in CPISP3 values between the 1980s and 1990s. From the 1990s to 2000s, the input of chemical fertilizer per hectare increased rapidly, whereas that of organic fertilizer decreased. The overall effect of these changes was to reduce the total nutrient input per hectare. In addition, the soils became increasingly acidic, although their soil organic matter (SOM) and total nitrogen (TN) concentrations remained relatively unchanged, and their AP and AK concentrations increased. Another potentially important factor is that shallow rotary tillage machines were widely adopted in China during the 2000s, which led to a reduction in the topsoil depth to only 12–20 cm. This decrease likely reduced the soil’s moisture and nutrient storage capacity (Shadrack et al. 2014). However, we attribute the bulk of the decline in CPISP3 values in China between the 1990s and 2000s to the reduced rates of organic fertilizer application. Experimental duration (yr) 0 5 10 15 20 25 30 35 CPISP (%) 0 20 40 60 80 100 Irrigated y =24.7+85.9e–0.1 x R 2 =0.61, P <0.01 Rainfed y =–0.4 x +52.5 R 2 =0.06, P =0.11 0 5 10 15 20 25 30 35 Continious maize y =35.7+38.1e –0.1 x R 2 =0.55, P <0.01 Maize/Soybean/Wheat y =62.6+39.4e –0.3 x R 2 =0.03, P= 0.26 A B Fig. 3 Contribution percentage of inherent soil productivity (CPISP) in mono-cropping system in Northwest (A) and Northeast (B) China. Bar indicates SD.